Allosteric Inhibition of Human Lymphoblast and Purified Porphobilinogen Deaminase by Protoporphyrinogen and Coproporphyrinogen

Home , Porphobilinogen

Allosteric inhibition of human lymphoblast and purified porphobilinogen deaminase by protoporphyrinogen and coproporphyrinogen. A possible mechanism for the acute attack of variegate porphyria.

P Meissner, … , P Adams, R Kirsch

J Clin Invest. 1993;91(4):1436-1444. https://doi.org/10.1172/JCI116348.

Research Article

Variegate porphyria (VP) is characterized by photocutaneous lesions and acute neuropsychiatric attacks. Decreased protoporphyrinogen oxidase activity results in accumulation of protoporphyrin (ogen) IX and coproporphyrin (ogen) III. During acute attacks delta-aminolevulinic acid and porphobilinogen also increase, suggesting that porphobilinogen deaminase (PBG-D) may be rate limiting. We have examined the effects of porphyrinogens accumulating in VP on PBG- D activity in Epstein-Barr virus-transformed lymphoblast sonicates from 12 VP and 12 control subjects. Protoporphyrinogen oxidase activity was decreased and protoporphyrin increased in VP lymphoblasts. PBG-D in control lymphoblasts obeyed Michaelis-Menten kinetics (Vmax 28.7 +/- 1.8 pmol/mg per h, Hill coefficient 0.83 +/- 0.07). VP sonicates yielded sigmoidal substrate-velocity curves that did not obey Michaelis-Menten kinetics. Vmax was decreased (21.2 +/- 2.0 pmol/mg per h) and the Hill coefficient was 1.78 +/- 0.17. Addition of protoporphyrinogen IX and coproporphyrinogen III to control sonicates yielded sigmoidal PBG-D substrate-velocity curves and decreased PBG-D Vmax. Addition of porphyrins or uroporphyrinogen III did not affect PBG-D activity. Removal of endogenous porphyrin (ogens) from VP sonicates restored normal PBG-D kinetics. Purified human erythrocyte PBG-D obeyed Michaelis-Menten kinetics (Vmax 249 +/- 36 nmol/mg per h, Km 8.9 +/- 1.5 microM, Hill coefficient 0.93 +/- 0.14). Addition of protoporphyrinogen yielded a sigmoidal curve with decreased Vmax. The Hill coefficient approached 4. These findings provide a rational explanation for the increased delta-aminolevulinic […]

Find the latest version: https://jci.me/116348/pdf Allosteric Inhibition of Human Lymphoblast and Purified Porphobilinogen Deaminase by Protoporphyrinogen and Coproporphyrinogen A Possible Mechanism for the Acute Attack of Variegate Porphyria Peter Meissner,* Paul Adams,* and Ralph Kirsch* MRC/UCT Liver Research Centre, *Departments ofMedicine and tChemical Pathology, University ofCape Town Medical School, Observatory 7925, South Africa

Abstract ity and a propensity to develop acute neuropsychiatric attacks with abdominal pain, vomiting, constipation, tachycardia, hy- Variegate porphyria (VP) is characterized by photocutaneous pertension, psychiatric symptoms, and, in the worst cases, sym- lesions and acute neuropsychiatric attacks. Decreased proto- metrical quadriplegia ( 1-3). porphyrinogen oxidase activity results in accumulation of pro- VP is associated with a defect of the penultimate heme syn- toporphyrin(ogen) IX and coproporphyrin(ogen) III. During thetic enzyme, protoporphyrinogen oxidase (4-8). This results acute attacks delta-aminolevulinic acid and porphobilinogen in accumulation of the distal heme synthetic intermediates also increase, suggesting that porphobilinogen deaminase protoporphyrin(ogen) IX and coproporphyrin(ogen) III. (PBG-D) may be rate limiting. We have examined the effects During acute attacks porphyrin(ogen) concentrations be- of porphyrinogens accumulating in VP on PBG-D activity in come even higher and, in addition, delta-aminolevulinic acid Epstein-Barr virus-transformed lymphoblast sonicates from (ALA) and porphobilinogen (PBG), proximal intermediates 12 VP and 12 control subjects. Protoporphyrinogen oxidase of the heme synthetic pathway, increase ( 1, 2, 9). activity was decreased and protoporphyrin increased in VP lym- Acute attacks also occur in two other forms of porphyria, phoblasts. PBG-D in control lymphoblasts obeyed Michaelis- acute intermittent porphyria and hereditary coproporphyria Menten kinetics (V.. 28.7±1.8 pmol/mg per h, Hill coeffi- (10). Although the site of the inherited enzyme defect varies, cient 0.83±0.07). VP sonicates yielded sigmoidal substrate-ve- PBG deaminase in acute intermittent porphyria (1 1, 12), co- locity curves that did not obey Michaelis-Menten kinetics. proporphyrinogen oxidase in hereditary coproporphyria ( 13, V,,,,, was decreased (21.2±2.0 pmol/mg per h) and the Hill 14), and protoporphyrinogen oxidase in VP, neuropsychiatric coefficient was 1.78±0.17. Addition of protoporphyrinogen IX attacks in all three acute porphyrias are associated with ele- and coproporphyrinogen III to control sonicates yielded sig- vated concentrations of ALA and PBG (10). moidal PBG-D substrate-velocity curves and decreased PBG- The elevated concentrations of ALA and PBG are easily D V,,,n. Addition of porphyrins or uroporphyrinogen III did explained in acute intermittent porphyria. However, this find- not affect PBG-D activity. Removal of endogenous ing does not tally with the known enzyme abnormality in VP porphyrin(ogens) from VP sonicates restored normal PBG-D and hereditary coproporphyria. The thesis that ALA and PBG kinetics. Purified human erythrocyte PBG-D obeyed Michae- are increased as a result of sequential damming up of products lis-Menten kinetics (V,,, 249±36 nmol/mg per h, K. proximal to a block in the distal portion of the heme synthetic 8.9±1.5 ,uM, Hill coefficient 0.93±0.14). Addition of protopor- pathway is unlikely since increased concentrations of ALA and phyrinogen yielded a sigmoidal curve with decreased V,,,n. The PBG are not found in porphyria cutanea tarda despite accumu- Hill coefficient approached 4. These findings provide a rational lation of large amounts of uroporphyrin(ogen) III, the inter- explanation for the increased delta-aminolevulinic acid and mediate immediately distal to PBG ( 10, 15-17). Similarly it is porphobilinogen during acute attacks of VP. (J. Clin. Invest. unlikely that subjects with VP and hereditary coproporphyria 1993. 91:1436-1444.) Key words: uroporphyrinogen-I syn- have a second inherited defect at the level of PBG deaminase thase - hydroxymethylbilane synthase * porphyria variegata- since the heme biosynthetic enzymes appear to be encoded on acute porphyria different chromosomes (9). In this study we examine the hypothesis that the obligate Introduction heme precursors that accumulate in VP (and in hereditary coproporphyria) inhibit PBG Variegate porphyria (VP),' an autosomal dominant error of deaminase activity and result in heme metabolism, is characterized rate-limiting effects. We have tested this hypothesis by per- by photocutaneous sensitiv- forming a series of kinetic analyses of PBG deaminase using EBV-transformed lymphocytes from normal and VP subjects. Address correspondence to Dr. P. Meissner, MRC/UCT Liver Re- Various porphyrins and porphyrinogens were assessed for their search Centre, Department of Medicine, K-Floor, Old Groote Schuur Hospital, Observatory 7925, South Africa. ability to alter kinetic behavior ofboth lymphoblast PBG deam- Receivedfor publication 4 May 1992 and in revisedform 19 Oc- inase and the purified enzyme. tober 1992. Sonicates of lymphoblasts derived from subjects with VP exhibit a significant decrease in PBG deaminase Vma activity 1. Abbreviations used in this paper: ALA, delta-aminolevulinic acid; and display abnormal sigmoidal PBG deaminase substrate-ve- PBG, porphobilinogen; VP, variegate porphyria. locity curves with kinetic features suggestive ofallosteric inhibition. These changes could be abolished by chromatography J. Clin. Invest. aimed at removing porphyrin(ogen)s from the PBG deami- © The American Society for Clinical Investigation, Inc. nase-containing fraction of VP lymphoblast sonicates and re- 0021-9738/93/04/1436/09 $2.00 produced by addition of protoporphyrinogen IX and copro- Volume 91, April 1993, 1436-1444 porphyrinogen III (but not uroporphyrinogen III or the por-

1436 P. Meissner, P. Adams, and R. Kirsch phyrins) to sonicates of lymphoblasts from normal subjects. 0.91±0.1-0.95±0.1 nmol/mg protein per h; PBG deaminase range of (Unless otherwise stated protoporphyrin[ogen] refers to the activity, 21.2±1.4-22.4±1.2 pmol/mg protein per h). Reproducibility series IX isomer, and coproporphyrin[ogen] and uropor- of the assays was verified by measuring the enzymes in triplicate in the phyrin[ogen] refer to the series III isomers.) Finally, kinetic same tissue batch on two separate occasions separated by 2 h. The changes similar to those found in VP lymphoblasts were found results differed nonsignificantly by < 14% (P = 5.853 X 10-'). when protoporphyrinogen was added to purified human PBG The effect of transformation on protoporphyrinogen oxidase and PBG deaminase Possible or effects of deaminase, suggesting that has a direct activity. stimulatory inhibitory protoporphyrinogen the transformation process on protoporphyrinogen oxidase and PBG on PBG deaminase effect activity. deaminase were assessed in five control cell lines by comparing activities of the two enzymes 3, 4, 5, 7, 14, 21, and 28 d after initiation ofthe Methods transformation process. Both protoporphyrinogen oxidase and PBG deaminase activity Subjects. 12 healthy control and 12 quiescent VP subjects were used as reached maximum (fourfold) increases in activity at 7 d. Thereafter, lymphocyte donors in this study. All were defined clinically and both enzyme activities dropped to a steady level ofapproximately twice by quantitative thin-layer chromatographic fluoroscanning of stool, the baseline value. All lymphoblast studies were performed during this urine, and plasma (18, 19). The VP subjects all had stool protopor- steady stage. phyrin concentrations > 200 nmol/g dry wt and coproporphyrin con- Evidence that the VP defect in protoporphyrinogen oxidase activity centrations > 50 nmol/g dry wt. Control subjects were asymptomatic, persists after transformation. We have previously shown that protopor- had no family history of porphyria, and exhibited normal porphyrin phyrinogen oxidase Vm. is decreased by 52% in VP-derived Iympho- biochemistry. All subjects gave informed consent for the use of their blasts when compared with normal controls (n = 27) (6). Additional blood, urine, and stool in this study. evidence for persistence of this defect was obtained by assaying proto- EB V-transformedlymphoblast preparation. EBV-transformed lyrm- porphyrinogen oxidase in fresh lymphocytes (500 ml blood) and in phoblasts were used in this study, since these cells could be derived lymphoblast cultures prepared from the same normal (n = 3) and VP from a relatively small amount of whole blood, they were easy to estab- subjects (n = 3). Both VP lymphocytes and lymphoblasts showed a lish in culture, and once established from each original specimen, they similar decrease in protoporphyrinogen oxidase activity (VP lympho- readily yielded the large numbers of cells required for kinetic studies cyte V,,. was 50% diminished, VP lymphoblast Va, was 55% dimin- (6). Most importantly, studies aimed at validating the lymphoblast ished). Km values were unaltered. system as a "model" of VP tissue allowed us to confidently assume the Measurement oflymphoblast porphyrins. Lymphoblasts were cen- "VP" nature of these cells (see below) . trifuged at 1,000 g for 15 min and washed twice in Hanks' Balanced 20 ml of blood from each subject was taken into heparin. Lympho- Salt Solution. After addition of 10 ml of 5% sulphuric acid in methanol cytes were isolated, cultured, and prepared for assay as previously de- to the final cellular pellet, the mixture was agitated vigorously and scribed (6). Medium containing EBV was prepared from semicon- heated in the dark at 55°C for 8 h. This ruptured the cells and esterified fluent cultures of the EBV-producing marmoset cell line B95/8 (20). any porphyrin present. Next, the mixture was centrifuged (2,000 g, 15 Protoporphyrinogen oxidase activity assay. Lymphoblast protopor- min), the supernatant solution was neutralized using a 17% (vol/vol) phyrinogen oxidase was assayed as previously described (6). Protopor- ammonia solution, and the porphyrin esters extracted and quantita- phyrinogen solution (produced by treating protoporphyrin with 4% tively analyzed by TLC (Silica gel-60 without fluorescent indicator; sodium amalgam) was added at five different concentrations (ranging Merck, Darmstadt, Germany) and fluoroscanning (TLD-100 scanning between 0.75 and 2.5 AM), to a lymphoblast sonicate, and reaction densitometer; Vitatron, Dieren, The Netherlands) as previously de- buffer (50 mM Tris-HCI pH 8.5, 1 mM EDTA, 3 mM dithiothreitol, scribed( 18, 19). and 1% Brij-35). The reaction was allowed to proceed in the dark for 1 Preparation and stability of porphyrinogens for use in kinetic as- h at 370C. Protoporphyrin concentrations were measured by direct says. Protoporphyrinogen, coproporphyrinogen, and uroporphyrino- fluorometry (model 204A fluorescence spectrophotometer; Perkin- gen (all porphyrins from Porphyrin Products, Inc., Logan, UT) for Elmer Cetus Instruments, Norwalk, CT). Nonenzymatic production addition to PBG deaminase assay systems (lymphoblast sonicates or of protoporphyrin was controlled for by measuring spontaneous proto- purified PBG deaminase) were prepared by reduction of the corre- porphyrin production under similar conditions. The rate of nonenzy- sponding porphyrin using 4% sodium amalgam that was freshly pre- matic oxidation of protoporphyrinogen was subtracted from the enzy- pared using well-described methods (28, 29). matic rate. Apparent V. and Km values for protoporphyrinogen oxi- A potential problem related to the known instability of porphyrino- dase were determined from double-reciprocal Lineweaver-Burk plots. gens: because testing of the hypothesis demanded that porphyrinogens All assays were performed in duplicate. be investigated, it was necessary to examine the stability of various PBG deaminase activity assay. PBG deaminase activity in lympho- porphyrinogens under assay conditions. The relative stability of proto-, blasts was measured by minimal adaptation of previously described copro-, and uroporphyrinogen under PBG deaminase assay conditions methods for measuring erythrocyte and lymphocyte PBG deaminase was studied as follows: 5 ml of a 50 jiM solution of the appropriate activity (21-25). porphyrin was reduced to the porphyrinogen using 4% sodium amal- Lymphoblasts were prepared for assay in reaction buffer (0.1 M gam. The reduced solution was filtered and the pH was adjusted to 7.5 Tris-HCl pH 8.2 and 0.1 mM dithiothreitol added fresh) as described with glacial acetic acid. 2 ml of this was added to 8 ml of PBG deami- above. PBG deaminase was released by sonication. Sonicates were cen- nase assay buffer and the solution incubated for 12 h at 37°C in a trifuged at 30,000 g for 45 min and the supernatants were used for shaking water bath in the dark for 2 h. A l-ml aliquot was removed assay. The reaction mixture contained 100 td of sample and 600 Jl of every 15 min and its fluorescence was measured in a fluorometer. After reaction buffer. PBG concentrations were selected to straddle the re- 2 h any porphyrinogen present in the remaining volume was reoxidized ported Km of PBG deaminase (22, 25-27). by adding 50 Al of a 0.005% (wt/vol) aqueous iodine solution. A small Evaluation oftransformed lymphoblastsforassayofprotoporphyrin- crystal of cysteine was added to decolorize the excess iodine (28). Oxi- ogen oxidase and PBG deaminase. The suitability of lymphoblasts as a dation of porphyrinogens as judged by reappearance of porphyrin fluo- model in which to study protoporphyrinogen oxidase and PBG deami- rescence was negligible over the first 30 min and very slight over the nase was assessed as follows. Four normal control lymphoblast lines next 90 min. Protoporphyrinogen was the least stable ( 14% reoxidation had protoporphyrinogen oxidase and PBG deaminase assayed at days after 120 min), and uroporphyrinogen was the most stable (4% reoxi- 30, 40, and 50 posttransformation to assess the variation in activity in dation after 120 min). It was concluded that the stability of the por- culture over a long period. Both enzymes showed remarkably constant phyrinogens under the proposed assay conditions was acceptable. activity over this period (protoporphyrinogen oxidase range of activity, Addition of porphyrin(ogens) in the kinetic assays. Protopor-

Inhibition ofPorphobilinogen Deaminase 1437 phyrin, coproporphyrin, and uroporphyrin were added to lymphoblast obtained from the Michaelis-Menten equation, is identical to K0.5 For sonicate incubation mixtures 5 min before the assay of PBG deaminase processes not following Michaelis-Menten behavior, the Km per se is activity. All additions were designed to achieve incubation mixture meaningless and an equivalent parameter, K0.5, is more informative. A porphyrin concentrations of 1, 5, and 10 MM. Protoporphyrinogen, Hill plot, log v/(V,|, - v) vs log [S] (where v is the initial velocity at the coproporphyrinogen, and uroporphyrinogen were added immediately given substrate concentration and [S] is the substrate concentration) after their preparation by reduction using 4% sodium amalgam. will have a slope approaching n where there is cooperativity. All com- In a preliminary study we compared two assay methods: direct fluo- parisons were assessed for statistical significance using the Student's t rometry of a 1 / 10 dilution of reaction mixture yielded a mean PBG test (35). deaminase Vmax of 26.2+4.1 pmol/mg protein per h whereas extraction and quantitative porphyrin TLC gave a mean V,., of 28.0±2.1 pmol/ mg protein per h. A comparison of these two methods showed a mar- Results ginally lower V. with a slightly larger scatter using the first method, although the two results were not statistically different (P = 2.300 Lymphoblast porphyrin measurements. Total porphyrin con- X 10-', n = 6). Furthermore, thin-layer chromatographic analysis in- centrations of VP lymphoblasts (1.41±0.21 pmol of total por- dicated similar quantities of copro- and protoporphyrin both before phyrin/ 106 cells, n = 7) was significantly more than that of and after assay, whereas the amount of uroporphyrin increased. It was normal control lymphoblast lines (1.10±0.18 pmol of total concluded that the use of fluorescence readings on diluted aliquots of reaction mixture was acceptable for the purposes of this study. porphyrin/ 106 cells, P = 1.935 X 10-3). The mean cell vol- Removal ofporphyrin(ogen)sfrom PBG deaminase in VPlympho- umes of these cells have been previously determined to be blasts. After sonication of VP lymphoblasts, Sephadex G25 (Pharma- 2.05±0.35 x 10-12 liters (36). This therefore represents a por- cia LKB Biotechnology Inc., Uppsala, Sweden) chromatography was phyrin concentration of - 2 MM. Roughly 90% was protopor- used to separate the low molecular weight endogenous porphyrin and the remainder coproporphyrin. phyrin(ogen)s from fractions containing PBG deaminase activity. When heme biosynthesis in lymphoblast cell lines was stim- Fractions containing PBG deaminase were used for subsequent kinetic ulated by growing in the presence of 0.6 mM ALA for 24 h studies. before harvesting according to previously described methods To validate the separation of porphyrins and PBG deaminase, a (21, 22), VP cell lines had significantly higher concentrations mixture of 0.5 mM uro-, copro-, and protoporphyrin was added to ofboth protoporphyrin (30% increased, P = 2.163 X 10-) and 10,000 g supernatants of control sonicates to ensure visi- lymphoblast coproporphyrin (14% increased, P = 3.254 x than con- ble porphyrin fluorescence in the resulting porphyrin containing frac- I0-3) tions eluted from the column. PBG deaminase activity was found to trol cells. elute in the G25 void whereas porphyrins eluted at 3.2-3.4 times the PBG deaminase kinetics in control and VP lymphoblasts. void volume. Lymphoblast sonicate PBG deaminase from 12 control sub- Next, supernatants (10,000 g) of sonicated VP lymphoblasts were jects gave a mean Vma, of 25.2±1.7 pmol/ mg protein per h and applied to a 1.5 X 10 cm Sephadex G25 column, equilibrated with a mean Km of 8.1±0.7 AM when evaluated by double-recipro- 0.1 M Tris-HCI buffer (pH 8.2), and eluted, with the same buffer, at a cal substrate vs velocity (Lineweaver-Burk) plots. Linearity of flow rate of 40 ml/h. Fractions containing PBG deaminase activity the Lineweaver-Burk plot and ofthe rate/ substrate concentra- were pooled, concentrated on Centricon-30 (Amicon, Beverly, MA) tion versus rate (Eadie) plot confirmed the adherence of PBG columns through a P30 membrane, and assayed for protein content deaminase in lymphoblasts derived from normal subjects before PBG deaminase assay. (control to the Michaelis-Menten rate law Purification ofPBG deaminase. PBG deaminase was purified from lymphoblasts) (Fig. human erythrocytes using a procedure adapted from two previously 1 A). Application of the Hill equation to the same data gave published methods (27, 30). Our purification data have recently been mean values for V.., of 28.7±1.8 pmol/mg protein per h and published (31). The clear single protein band obtained with silver KO5 of 8.5±0.8 MM that were not significantly different from staining was estimated to have a molecular mass of41,200 D (±2,300) those obtained using the Michaelis-Menten equation. The Hill and the final specific activity of 241 nmol uroporphyrin formed/mg coefficient was 0.83. per h represented - 7, 100-fold purification. Identical experiments in sonicates of lymphoblasts from 12 Addition ofporphyrinogen and porphyrin to purified PBG deami- VP subjects resulted in a 26% decrease in Vm. (21.2±2.0 vs nase. Substrate-velocity data were obtained over a 1.25-20MOM concen- 28.7±1.8 pmol/mg protein per h; P = 6.119 X 10-4) and tration range of PBG at protoporphyrinogen concentrations of 0.01, yielded sigmoidal substrate-velocity curves (Fig. 1 B) with data 0.05, 0.1, 0.5, 1, 5, and 10 M. The effect ofadded protoporphyrin was that did not conform to assessed at 0.1, 1, and 10 MM concentrations. All enzyme assays were Michaelis-Menten kinetics. Nonhy- performed in duplicate. perbolic behavior was confirmed by failure of the data to pro- Kinetic analysis and statistics. Data obtained from assays of PBG duce a straight line with either the Lineweaver-Burk or Eadie deaminase reaction velocity at different substrate concentrations using plots. Thus Vm,,, and Km could not be assessed by this method. lymphoblasts were initially treated by application of Michaelis-Men- The Hill equation, however, did allow these determinations ten equation (32) to obtain VP Vm.,, and Km. Data which did not and was therefore used. conform to the Michaelis-Menten rate law were examined using the Addition of porphyrinogens to control lymphoblast soni- Hill equation (33). A nonlinear least-squares grid search procedure cates. Addition of protoporphyrinogen at 1, 5, and 10MM con- computed the following parameters from the Hill equation, which were centrations altered the kinetic behavior of control lymphoblast used for comparative purposes: Vm,,, Ko.5, and the Hill coefficient, n. PBG deaminase (Fig. 2 A and Table I). activity was re- Vn, is the maximal catalytic rate (velocity) for PBG deaminase, K0.5 is V., the substrate concentration at which half maximal velocity is achieved. duced and the substrate-velocity plots became sigmoidal, re- The Hill coefficient describes the extent of cooperativity for ligand- sembling those observed in VP lymphoblasts. binding processes or the extent to which multiple ligand-binding sites Compared with the values obtained in the absence ofadded interact. The coefficient n, is always less than the actual number of porphyrinogen, Vma. was decreased by an average of 41% (P ligand-binding sites (34). Thus, for single-site substrate-binding pro- = 7.004 X 10-4) whereas the K05 was unaltered (P = 3.611 cesses displaying perfect hyperbolic Michaelis-Menten behavior, the X 10 1). Vm,, tended to decrease as the concentration ofproto- Hill coefficients will approximate to 1. For hyperbolic behavior, Km as porphyrinogen increased, but the relationship did not achieve 1438 P. Meissner, P. Adams, and R. Kirsch A The distinctive sigmoidal substrate-velocity plot was once x10 again observed, as were the trends in V. and K.5 (Fig. 2 B and Table 1). Longer periods of preincubation in the presence of 1.6 coproporphyrinogen showed no further changes to those measured after 5 min of preincubation. E 1.2 Addition of uroporphyrinogen and porphyrins to control E lymphoblast sonicates. In contrast to the striking effects noted 0 E on the addition of proto- and coproporphyrinogen, uropor- ,,,w 0.8 phyrinogen failed to influence the kinetic behavior of control !R lymphoblast PBG deaminase. There were no significant differ- ences in V,,,., K0.5, the Hill coefficient (P > 0.05 in all cases), 0.4 or in the shape of the substrate-velocity plot (Table 1). Similarly, the addition of 1, 5, or 10 MM proto-, copro-, or PBG 0 uroporphyrin had no effect on the kinetic behavior of 0 0.4 0.8 1.2 1.6 x10 deaminase in 12 control lymphoblast sonicates (Table I). In- [PBG] (pM) deed, preincubating lymphoblast preparations with 10 AM proto-, copro-, and uroporphyrin for up to 1 h before assaying B PBG deaminase also did not alter PBG deaminase behavior. X10 Addition ofporphyrinogens to VP lymphoblast sonicates. 1.6 Addition of proto- and coproporphyrinogen to VP lymphoblast sonicates resulted in an exaggeration of the kinetic changes already present. Once again, protoporphyrinogen ap- _ 1.2 0) E 0 E X10 A 0.8 w 1.6

0.4 _ 1.2 E 0 E 0 a 0.8 0 0.4 0.8 1.2 1.6 x1O w I- [PBG] ir Figure 1. Control and VP lymphoblast PBG deaminase kinetics. (A) 0.4 Control lymphoblast sonicates (n = 12) demonstrated a hyperbolic substrate-velocity plot. Michaelis-Menten kinetics were obeyed as il- lustrated by the Lineweaver-Burk (inset 1) and Eadie plots (inset 2). 0 (B) In contrast VP lymphoblast PBG deaminase demonstrated a sig- 0 0.4 0.8 1.2 1.6 xlO moidal substrate-velocity plot that could be evaluated using the Hill [PBG] equation. Inset shows the Hill plot. B xlO statistical significance. The K.5 obtained in the presence of 10 ,MM protoporphyrinogen was greater than that obtained with 5 1.6 MAM protoporphyrinogen, which was in turn greater than that in the presence of 1 uM protoporphyrinogen, but none achieved significance. 1.2 As the concentration of protoporphyrinogen was increased EE' 0 over a 10-fold range, the Hill coefficient increased to approach E 4 in 1 of - 0.8 a value of the limit (Table ). Longer periods preincuba- w tion in the presence of protoporphyrinogen appeared to exert no further effect on the kinetic behavior ofcontrol lymphoblast 0.4 PBG deaminase. Addition of 1, 5, and 10 ,M concentrations of coproporphyrinogen had a similar but less impressive effect on control 0 lymphoblast PBG deaminase. As coproporphyrinogen concen- 0 0.4 0.8 1.2 1.6 x10 trations were increased over a 10-fold range, V.. decreased by [PBG] an average of 27% (P = 7.097 X 10-4) and the Hill coefficient Figure 2. Substrate-velocity plots of PBG deaminase in control lym- appeared to approach 3 (in the limit). The mean of the K0.5 phoblast sonicates in the absence (dashed line) and presence of (A) values between the lymphoblasts with additions and those 1 (o), 5 (o), and 10 MM (A) protoporphyrinogen and (B) 1, 5, and without were not significantly different (P = 2.474 X 10-'). 10,gM coproporphyrinogen.

Inhibition ofPorphobilinogen Deaminase 1439 Table I. Control PBG Deaminase Kinetic Parameters (as A plot of Vm.. and the Hill coefficient versus protoporphy- Obtainedfrom the Hill Equation) in the Absence and Presence of rinogen concentration (Fig. 4) shows a single exponential 1, 5, and 10 AM Protoporphyrinogen, Coproporphyrinogen, growth relationship in the case of the Hill coefficient and a Proto-, Copro-, and Uroporphyrin single exponential decay for V,,.. Single exponential analysis Hill reveals that the protoporphyrinogen concentration at which Control lymphoblasts V,,,.,, K0.5 coefficient the Hill coefficient is half of its estimated maximum is 0.2 ,M, and the concentration at which Vm,,, is half its estimated maxi- pmol/mg per h AM mum inhibition is 0.19 MM. No additions 28.7±1.8 8.5±0.8 0.83±0.07 Addition ofprotoporphyrin at 0. 1, 1, and 10MM concentra- +1 ItM proto'gen 17.7±1.3 7.7±1.1 2.89±0.43 tions had no effect on pure PBG deaminase kinetics. +5 AM proto'gen 16.4±1.7 8.9±1.1 3.05±0.47 Sephadex G25 chromatography. PBG deaminase kinetics +10 MuM proto'gen 16.1±1.6 10.9±2.1 3.61±0.45 ofcontrol lymphoblast preparations were not changed by Seph- +1 MuM copro'gen 21.8±1.8 8.5±2.3 1.67±0.51 adex G25 chromatography. In striking contrast, after Sephadex +5 AM copro'gen 21.5±2.2 9.8±1.3 1.96±0.33 G25 chromatography, VP lymphoblast PBG deaminase sub- +10 AM copro'gen 19.5±1.9 10.3±1.4 2.51±0.38 strate velocity data were identical to those of control lympho- +1 AM uro'gen 30.1±1.5 7.8±1.2 0.79±0.26 blasts (Fig. 5). +5 MM uro'gen 28.7±2.4 8.7± 1.0 0.93±0.11 +10 MM uro'gen 29.0±1.6 8.4±1.1 0.89±0.13 Discussion +1 MuM proto 28.4±2.9 8.4± 1.1 0.89±0.10 +5 MM proto 28.4±1.7 8.3±0.9 0.90±0.09 Much of the evidence favoring our hypothesis that heme pre- +10MM proto 29.3±1.6 8.0±1.0 0.87±0.10 cursors that accumulate in VP inhibit PBG deaminase activity +1 MM copro 30.2±2.1 8.8±1.9 0.86±0.10 depends on the validity of the EBV-transformed lymphoblast +5 MM copro 31.2±1.3 8.9±1.2 0.89±0.05 system. Two lines of evidence suggest that the VP defect per- +10MM copro 30.8±1.6 9.5±0.9 0.88±0.07 sists in this system. First, we have confirmed our previously + 1 MM uro 28.8±1.3 9.0±1.0 0.95±0.11 published data that protoporphyrinogen oxidase activity in VP +5 MM uro 34.5±3.3 8.5±1.3 0.78±0.16 lymphoblasts is reduced by roughly 50% (6). Second, and vital +10 AM uro 30.9±2.0 8.7±1.3 0.82±0.08 to our hypothesis, there was a significant increase in endogenous protoporphyrin concentration in our VP lymphoblast Proto'gen, protoporphyrinogen; copro'gen, coproporhyrinogen; cultures. Indeed when porphyrin production was stimulated by uro'gen, uroporphyrinogen; proto, protoporphyrin; copro, copropor- addition of ALA to the culture medium both protoporphyrin; uro, uroporphyrin. phyrin(ogen) and coproporphyrin(ogen) concentrations in VP lymphoblasts exceeded those of similarly treated cells derived from control subjects. Although diminished protoporphyrinogen oxidase activity peared to have a greater effect than coproporphyrinogen. Vm.. has previously been demonstrated in fibroblasts from VP sub- did not decrease to levels lower than those observed when 10 jects, protoporphyrin(ogen) concentration in that system was MM protoporphyrinogen was added to control lymphoblast not increased even after the addition of ALA (4). The elevated sonicates. Similarly, the Hill coefficients never tended to a protoporphyrin concentrations found in our study suggest that value > 4. in transformed lymphoblasts, the flux through the pathway As in the control lymphoblasts, uroporphyrinogen failed to may be high enough for the defect in protoporphyrinogen oxi- exert any additional effect on VP PBG deaminase kinetic be- dase to become evident. havior (Table II). Furthermore, the addition of 1, 5, or 10 AM proto-, copro-, or uroporphyrin had no additional effect on the already anomalous kinetic behavior of PBG deaminase in 12 VP lymphoblast sonicates. Table I. VP PBG Deaminase Kinetic Parameters Kinetics of PBG deaminase. in the Absence and Presence of 1, 5, and 10 AM Proto-, purified Substrate-velocity Copro-, and data ofpure PBG deaminase yielded a hyperbolic curve (Fig. 3, Uroporphyrinogen dotted line and Table with a III) V,_ of 249+36 nmol/mg per Hill h, a Ko5 of 8.9±1.5 MM, and a Hill coefficient of 0.93±0.14. VP lymphoblasts V.x K,, coefficient Lineweaver-Burk and Eadie plots were linear. Addition of protoporphyrinogen over a 1,000-fold range of pmol/mg per h AM concentrations (0.01-10MIM) transformed the hyperbolic sub- No additions 21.2±2.0 7.4±0.7 1.78±0.17 strate-velocity curve obtained in the absence ofprotoporphyrin- +1 AM proto'gen 17.7±1.8 7.2±1.0 2.25±0.40 ogen to sigmoidal curves similar to those displayed by impure +5 MM proto'gen 17.3±1.5 8.5±1.2 2.64±0.49 PBG deaminase in the presence of 1, 5, and 10MM proto-, and +10 AM proto'gen 16.6±1.3 8.7±1.1 3.22±0.42 coproporphyrinogen (Fig. 3, solid lines and Table III). There +1 AM copro'gen 19.0±1.5 8.5±1.1 1.88±0.31 was an inverse relationship between Vm,, and the protopor- +5 MM copro'gen 18.5±2.0 7.9±2.2 2.01±0.33 phyrinogen concentration. Vm,, in the presence of 10 MM pro- +10MM copro'gen 18.3± 1.8 8.9± 1.4 2.05±0.31 toporphyrinogen, 169±28 nmol/mg per h, was 32% lower than +1 AM uro'gen 20.4±1.4 7.5± 1.6 1.75±0.19 that in the absence of protoporphyrinogen. The Hill coeffi- +5 MM uro'gen 19.5±1.3 9.4±1.8 1.75±0.18 cients appeared to approach 4 in the limit, and Ko 5 was essen- +10 AM uro'gen 21.3±2.2 7.6±0.9 1.83±0.20 tially unaltered.

1440 P. Meissner, P. Adams, and R. Kirsch x1021 reports of nonMichaelian behavior from one group (26, 27, 40). In our study, the impure preparations of PBG deaminase 1.6F from control lymphoblasts exhibited classic Michaelis-Menten behavior. This was borne out by the linearity of the Line- weaver-Burk plot ( I /rate vs 1 /[substrate] ) and the Eadie plot CD 1.2 (rate/[substrate] vs rate) (Fig. 1 A). E The Hill equation (33) can also be used for determining 0 E kinetic indexes. Under conditions where there is a hyperbolic substrate-velocity relationship, the Hill coefficient approaches 1 and should yield V. and K0.5 values equivalent to the Mi- chaelis-Menten V. and Km. This was the case in control lym- 0.4 phoblast PBG deaminase. The Hill equation is also used to analyze the number of ligand-binding sites and their interde- pendence in macromolecular species such as proteins. It has been used in this regard to directly demonstrate variable li- 0 0 0.4 0.8 1.2 1.6 x10 gand-binding stoichiometries in the interaction ofpyridine and N-methyl pyridinium cation with ferriprotoporphyrin IX (41 ). [PBG] A Hill coefficient close to 1 implies that the enzyme has only Figure 3. Addition of 0.01-10 MM protoporphyrinogen to pure PBG one binding site or that multiple sites, if present, are identical deaminase resulted in sigmoidal substrate-velocity plots. Dotted line and that the binding at any site is independent of the occupa- represents substrate-velocity plot without additions. tion of other sites. It is generally accepted that PBG deaminase exists in monomeric form and that there is only one substrate (ligand)-binding site (22, 30). This is consistent with our value Substrate-velocity data for PBG deaminase in lymphoblast for control lymphoblasts. sonicates were obtained by measuring the rate of formation of The substrate-velocity data for VPlymphoblast PBG deami- uroporphyrinogen at five substrate concentrations. Ideally nase does not fit the Michaelis-Menten equation (Fig. 1 B). these concentrations should fall both above and below the Km The most striking difference was a sigmoidal substrate-velocity for the enzyme (33) so as to cover a wide range around the plot. Fortunately, the data fitted the Hill equation well. Using Vm,.. The literature reports Km values ranging between 6 and this equation, the mean V.. value of PBG deaminase in VP 22 AM (22, 25-27, 30) and our data indicated a Km (or the cells was 21.2±2.0 pmol/mg protein per h, a significant 26% equivalent K0.5 using Hill analysis) of - 8 AM for lymphoblast decrease over that of control cells. Comparison of the K0.5 val- PBG deaminase originating from healthy control subjects. ues obtained in the VP and control groups showed no signifi- Thus the choice of PBG concentrations of 1.25, 2.5, 5, 10, and cant alterations. This implies that there was no major differ- 20 MM is appropriate, since these substrate concentrations ence in the thermodynamics of ligand binding by PBG deami- cover a range from 13 to 73% of V,. nase in the two groups. A sigmoidal substrate-velocity curve Analysis of the data obtained in control lymphoblasts by implies the presence ofmore than one cooperative ligand-bind- the Michaelis-Menten equation yielded a V,, of 25.2±1.7 ing site. The substrate-velocity plot of VP lymphoblast PBG pmol/mg per h and a Km of 8.1±0.7 AM. There are only two deaminase yielded a Hill coefficient of 1.78±0.17. This sug- previous reports of the measurement of PBG deaminase activ- gests that, in addition to the reduction in V.. VP lymphoblast, ity in lymphoblasts. One (25) reports an activity of 11 pmol/ PBG deaminase might exist in a form displaying multiple coop- mg protein per h for human lymphoblast lines, but it is unclear erative binding sites. This might reflect either the unmasking of whether these were EBV-transformed lymphoblast lines. The preexisting additional sites or the formation of a cooperative other study (22) found values of 40 and 68 pmol/mg protein multimeric structure with decreased catalytic ability. per h for PHA- or pokeweed-mitogen-stimulated lymphocytes and EBV-transformed lymphoblasts, respectively. It must be noted that reported erythrocyte PBG deaminase activities show a large range, which underscores the need to establish control Table III. Addition of Various Concentrations of Protoporphyrinogen to Pure PBG Deaminase Resulted in Altered values for each individual laboratory and method. Our control Kinetics as Determined by the Hill Equation lymphoblast results appear to be of the same order as those in other reports and were thus considered valid for the purposes of Purified control Hill this study. PBG deaminase V. Koj coefficient Comparison of the indexes of kinetic behavior of VP lym- pmol/mg per h AM phoblast PBG deaminase with that of control lymphoblasts yielded the unprecedented observation that although control No additions 249±36 8.9±1.5 0.93±0.14 PBG deaminase exhibited a hyperbolic substrate-velocity rela- +0.0 1 AM proto'gen 246±37 8.9±1.0 0.96±0.09 tionship suggesting enzyme substrate interaction at a single li- +0.05 AM proto'gen 235±42 9.5±1.5 1.19±0.14 gand-binding site, VP PBG deaminase did not (Fig. 1). +0.10MAM proto'gen 221±40 9.8±1.1 1.58±0.20 Although the mechanism of action of PBG deaminase re- +0.50MgM proto'gen 190±48 9.1±1.9 2.27±0.16 mains a subject of research activity and debate (37-39), it is + 1.00 AM proto'gen 186±46 9.7± 1.6 2.66±0.16 generally assumed that in most cases the appearance and rate +5.00MuM proto'gen 178±39 9.7±1.4 3.02±0.11 of production of uroporphyrinogen from PBG by the enzyme +10.0MAM proto'gen 169±28 9.7±1.6 3.49±0.19 follows the Michaelis-Menten rate law. There are, however,

Inhibition ofPorphobilinogen Deaminase 1441 Vmax (nmol/mg/h) HILL COEFFICIENT

0.5 0 1 2 3 4 5 6 7 8 9 10 PROTO'GEN CONCENTRATION Figure 4. Relationship between added protoporphyrino- -Vmax - Hill coefficient gen and Vmax and the Hill coefficient.

Addition of proto- and coproporphyrinogen to control lym- were less marked than those produced by protoporphyrinogen. phoblast preparations at concentrations bracketing those ob- For 1 and 5 ,uM coproporphyrinogen the Hill coefficient ap- served in VP lymphoblasts resulted in PBG deaminase kinetics proached 2 and for 10,gM coproporphyrinogen approached 3. that closely resembled those seen in VP lymphoblasts. More Thus coproporphyrinogen may be a less "potent" negative ef- specifically, addition of protoporphyrinogen (Fig. 2 A), de- fector. Addition of uroporphyrinogen or of porphyrin did not creased Vmax and changed the substrate-velocity plot to a sig- influence control lymphoblast PBG deaminase kinetics. moidal form. Vm. decreased most after addition of 10 uM The similarity between the behavior of PBG deaminase in protoporphyrinogen and least after 1 uM protoporphyrinogen VP lymphoblasts and those observed in normal lymphoblasts (Table I) but this decrease in activity did not appear to be after addition ofprotoporphyrinogen and coproporphyrinogen directly proportional to the amount of porphyrinogen added. suggest that the sigmoidal curve of VP cell PBG deaminase This might reflect a maximal effect at a concentration of 5 or might be related to intracellular accumulation of these heme 10,gM or may result from the use of impure PBG deaminase. intermediates. If this were so, addition ofthese porphyrinogens Addition of protoporphyrinogen increased the Hill coeffi- to VP lymphoblast PBG deaminase preparations should en- cient to a maximum of 3.61. This would suggest that, in re- hance the changes already present. Our finding that additional sponse to addition of protoporphyrinogen, PBG deaminase ex- protoporphyrinogen and coproporphyrinogen, but not uropor- hibits up to four ligand-binding sites because of aggregation of phyrinogen, proto-, copro-, or uroporphyrin, decreased V. enzyme protein or the unmasking ofhitherto unsuspected bind- and increased the Hill coefficient to levels not dissimilar to ing sites. Addition of coproporphyrinogen to control lympho- those obtained when 10 uM concentrations of proto- and co- blast preparations produced effects similar to those observed proporphyrinogen were added to normal lymphoblast soni- after addition of protoporphyrinogen. The changes in shape of cates is in keeping with this thesis. the substrate-velocity curve and the diminution Vma, activity To exclude any confounding effect associated with the use of less pure lymphoblast sonicates, a limited series of experiments were performed using purified human erythrocyte PBG x10 deaminase. Our purification procedure, judged by silver staining of SDS-PAGE, yielded homogeneous PBG deaminase. The 1.6 final specific activity, 241 nmol uroporphyrin formed/mg per h, was intermediate between the 2300 nmol/mg per h (25) .- E' 1.2 and the 68 nmol/mg per h (27) reported by the two groups on whose techniques our method was - 0 based. Our 7,100-fold E purification lies between the 3,400-fold (27) and 42,000-fold w 0.8 (25) reported by the two methods. The product ran on SDS- PAGE as a single unit of molecular weight 41,200 (±2,300). Purified PBG deaminase, in the absence of added porphy- 0.4 rinogen, behaved as a Michaelian enzyme with a single substrate-binding site. Addition of protoporphyrinogen to purified 0 PBG deaminase once again resulted in a decreased Vma and a 0 0.4 0.8 1.2 1.6 X1 0 sigmoidal substrate-velocity curve. Once again, the Hill coeffi- [PBG] cient approached 4 (Fig. 4 and Table III). Both Vm. and the Figure 5. VP lymphoblast PBG deaminase substrate-velocity plots Hill coefficient were maximally affected at 10 uM protopor- before (A&) and after (A) Sephadex G-25 chromatography. The kinetic phyrinogen and tenuously affected at 0.01,M. Single exponen- parameters were as follows: pre- and (post)-Sephadex G-25 chroma- tial analysis of both Vm,. and the Hill coefficient against proto- tography Vmax, 19.9±1.9 (30.8+1.6); KO.5, 7.9±1.2 (6.7±2.9); Hill porphyrinogen concentration (Fig. 4) revealed that the con- coefficient, 1.78±0.21 (0.82±0.33). centrations at which the effects were half "maximal" were

1442 P. Meissner, P. Adams, and R. Kirsch almost identical. This strongly suggests that a single mecha- porphyria, the intermediates accumulating as a result of the nism is responsible for affecting both Vm., and the degree of inherited enzyme defect may partially inhibit PBG deaminase. cooperativity. Addition of protoporphyrin had no effect on the The finding that uroporphyrinogen exerted no influence on kinetics of pure PBG deaminase. These findings using purified PBG deaminase is in keeping with our hypothesis, since por- erythrocyte enzyme are so similar to those obtained using our phyria cutanea tarda is a nonacute porphyria in which concen- lymphoblast sonicates that they suggest that both erythroid and trations of ALA and PBG are not elevated despite greatly in- nonerythroid forms of PBG deaminase are inhibited by proto- creased uroporphyrinogen concentrations. porphyrinogen. If the kinetic abnormalities observed in VP lymphoblasts are indeed due to intracellular accumulation of proto- and coproporphyrinogen, removal of these porphyrinogens from VP We thank R. Schmid, R. Hift, W. Gevers, M. Berman, P. Berman, and lymphoblasts should restore normal PBG deaminase kinetic M. Moore for advice; E. Sturrock, A. Corrigall, J. Sutherland, B. Da- behavior. After establishing that fractions containing PBG de- vidson, and A. Damon for technical assistance; and I. Batho for manu- aminase activity and those containing porphyrin(ogen)s could script preparation. be separated by Sephadex G25 chromatography, this procedure was used to separate endogenous heme intermediates References from VP lymphoblast fractions exhibiting reduced PBG deaminase activity. Single passage ofVP lymphoblast cytosolic prepa- 1. Eales, L., R. S. Day, and G. H. Blekkenhorst. 1980. The clinical and bio- rations down a Sephadex G25 column resulted in restoration chemical features of variegate porphyria: an analysis of 300 cases studied at Groote Schuur Hospital, Cape Town. Int. J. Biochem. 12:837-853. of "normal" kinetic behavior to PBG deaminase. Although it is 2. Day, R. S. 1986. Variegate porphyria. Semin. Dermatol. 5:138-154. possible that molecular sieving removed other factor(s) respon- 3. Meissner, P. N., D. M. Meissner, E. D. Sturrock, B. P. Davidson, and R. E. sible for alteration of PBG deaminase behavior, these results Kirsch. 1987. Porphyria-the UCT experience. S. Afr. Med. J. 72:755-76 1. 4. Brenner, D. A., and J. R. Bloomer. 1980. The enzymatic defect in variegate are consistent with our hypothesis. porphyria. N. Engl. J. Med. 302:765-769. Finally, if accumulation of proto- and coproporphyrinogen 5. Deybach, J. C., H. De Verneuil, and Y. Nordmann. 1981. The inherited affect PBG deaminase in VP, urinary ALA and PBG concen- enzymatic defect in porphyria variegata. Hum. Genet. 58:425-428. 6. Meissner, P. N., R. S. Day, M. R. Moore, P. B. Disler, and E. Harley. 1986. trations should be marginally elevated, even in quiescent VP. Protoporphyrinogen oxidase and porphobilinogen deaminase in variegate por- ALA and PBG levels have traditionally been considered to be phyria. Eur. J. Clin. Invest. 16:257-261. "normal" in quiescent VP. We have measured urinary concen- 7. Viljoen, D. J., R. Cummins, J. Alexopoulos, and S. Kramer. 1983. Proto- porphyrinogen oxidase and ferrochelatase in porphyria variegata. Eur. J. Clin. trations of these precursors in 221 patients with quiescent VP Invest. 13:283-287. and 93 controls. Urinary ALA and PBG concentrations in qui- 8. Boyle, J., Hordovatzi, G. G. Thompson, and M. R. Moore. 1986. The use escent VP subjects were 32.4±26.3 and 15.5±23.6 ,umol/ 10 of leucocyte protoporphyrinogen oxidase activity in screening a family with varie- mmol creatinine, 2.5 and 4.6 times higher, respectively, than gate porphyria. Biochem. Soc. Trans. 14:153-154. 9. Nordmann, Y., J. C. Deybach. 1990. Human hereditary porphyrias. In those of control subjects (ALA 12.85±5.7, PBG 3.4±2.3) (P Biosynthesis of Heme and Chlorophylls. H. A. Dailey, editor. McGraw-Hill, New < 0.001). This increase in urinary ALA and PBG excretion York. 491-542. suggests that the decrease in PBG deaminase Vm,,, is sufficient 10. Kappas, A., S. Sassa, and K. E. Anderson. In The Porphyrias. J. B. Stan- bury, J. B. Wyngaarden, and D. S. Frederickson, editors. McGraw-Hill, New to perturb the heme synthetic processes even in quiescent VP. York. 1301-1384. The mechanism by which protoporphyrinogen and to a 11. Miyagi, K. 1970. Deficiency of hepatic porphobilinogen deaminase in lesser degree coproporphyrinogen affect PBG deaminase is not acute intermittent porphyria. J. Kyushu Hematol. Soc. 20:190-203. the these intermediates 12. Strand, L. J., U. A. Meyer, B. F. Felsher, A. G. Redeker, and H. S. Marver. known. Our data suggest that ability of 1972. Decreased red cell uroporphyrin I synthetase activity in intermittent acute to inhibit PBG deaminase may be related to the number of porphyria. J. Clin. Invest. 51:2530-2536. carboxyl groups on the porphyrinogen. Uroporphyrinogen had 13. Elder, G. H., J. 0. Evans, N. Thomas, R. Cox, M. J. Brodie, M. R. Moore, no effect, whereas the effect of coproporphyrinogen was less and A. Goldberg. 1976. The primary enzyme defect in hereditary coproporphyria. Lancet. 2:1217-1219. than that of protoporphyrinogen. Thus steric and/or hydro- 14. Grandchamp, B., and Y. Nordmann. 1977. Decreased lymphocyte copro- phobic effects may be important. Secondly, the lack of influ- porphyrinogen III oxidase activity in hereditary coproporphyria. Biochem. ence of porphyrins on PBG deaminase may be caused by the Biophys. Res. Commun. 74:1089. 15. Pimstone, N. R. 1982. Porphyria cutanea tarda. Semin. Liver Dis. 2:132- strongly aromatic nature of these compounds. This is in con- 142. trast to the unconjugated macrocyclic ring system of the por- 16. Moore, M. R., K. E. L. McColl, C. Rimington, and A. Goldberg. 1987. phyrinogens. In the porphyrins the conjugated pi-electron sys- Disorders of Haem Metabolism. Plenum Publishing Corp., New York. below the whole 17. Brodie, M. J., M. R. Moore, G. G. Thompson, B. C. Campbell, and A. tems lie above and tetrapyrrolic macrocycle Goldberg. 1977. Is porphobilinogen deaminase activity a secondary control mech- and may shield the porphyrin from potential interaction with anism in haem biosynthesis in humans? Biochem. Soc. Trans. 5:1466-1468. the enzyme molecule. The porphyrinogens on the other hand 18. Day, R. S., N. R. Pimstone, and L. Eales. 1978. The diagnostic value of are aliphatic and the aromaticity is confined to the individual blood plasma porphyrin methyl ester profiles produced by quantitative TLC. Int. J. Biochem. 9:897-904. pyrrolic "corners" (42). This renders the interpyrrolic methyl- 19. Day, R. S., R. E. de Salamanca, and L. Eales. 1978. Quantitation of red ene groups more accessible. cell porphyrins by fluorescence scanning after thin layer chromatography. Clin. Our data provide several lines of evidence to support our Chim. Acta. 89:25-33. exert 20. Miller, G., and M. Lipman. 1973. Release of infectious Epstein-Barr virus hypothesis that proto- and coproporphyrinogen negative by transformed marmoset leucocytes. Proc. Natl. Acad. Sci. USA. 70:190-194. allosteric effects on PBG deaminase. This may account for the 21. Sassa, S., S. Granick, D. R. Bickers, H. L. Bradlow, and A. Kappas. 1974. increases of ALA and PBG seen when heme synthesis is de- Studies in porphyria III. A microassay for uroporphyrinogen I synthetase, one of during the acute porphyric attack. Our study offers an three abnormal enzyme activities in acute intermittent porphyria and its applica- pressed tion to the study of the genetics of this disease. Proc. Natl. Acad. Sci. USA. explanation for the apparent defect in PBG deaminase in each 7 1:732-736. of the acute porphyrias. In acute intermittent porphyria, PBG 22. Sassa, S., G. L. Zalar, and A. Kappas. 1978. Studies in porphyria VII. deaminase deficiency is inherited. In VP and hereditary copro- Induction of uroporphyrinogen I synthetase and expression of the gene defect of Inhibition ofPorphobilinogen Deaminase 1443 acute intermittent porphyria in mitogen stimulated lymphocytes. J. Clin. Invest. Analysis. H. U. Bermeyer, editor. Verlag Chemie, Weinheim, Germany/New 61:499-508. York. 13-86. 23. Piepkorn, M. W., Hammernyik, and R. F. Labbe. 1978. Modifiederythro- 33. Van Holde, H. E. 1971. Physical Biochemistry. Prentice Hall, Englewood cyte uroporphyrinogen I synthase assay and its clinical interpretation. Clin. Cliffs, NJ. Chem. 24:1751-1754. 34. Hammes, G. G. 1982. Enzyme Catalysis and Regulation. Academic Press, 24. Ford, R. E., C. On, and R. D. Ellefson. 1980. Assay forerythrocyte uropor- London. phyrinogen I synthase activity with porphobilinogen as substrate. Clin. Chem. 35. Bailey, N. T. J. 1981. Statistical Methods in Biology. Hodder and 26:1182-1185. Stoughton, London. p. 43. 25. Anderson, P. M., and R. J. Desnick. 1982. Porphobilinogen deaminase: 36. Steyn, L. M., and E. H. Harley. 1985. Intracellular activity of methods and principles of the enzymatic assay. Enzyme (Basel). 28:146-157. HPRTc.,. Town,: Purine uptake and growth of cultured cells in selective media. J. 26. Sancovich, H. A., A. M. del C Batlle, and M. Grinstein. 1969. Porphyrin Inherited Metab. Dis. 8:198-203. synthesis VI. Separation and purification of porphobilinogen deaminase and uro- 37. Jordan, P. M., and M. J. Warren. 1987. Evidence for a dipyrromethane porphyrinogen isomerase from cow liver. Porphobilinogenase an allosteric en- cofactor at the catalytic site of E. coli porphobilinogen deaminase. FEBS Fed. zyme. Biochim. Biophys. Acta. 191:130-143. Eur. Biochem. Soc., Lett. 225:87-92. 27. Fumagalli, S. A., M. L. Kottler, M. V. Rossetti, and A. M. del C Batlle. 38. Battersby, A. R. 1988. Biosynthesis of the pigments of life. J. Nat. Prod. 1985. Human red cell porphobilinogen deaminase. A simpler method of purifica- 5 1:629-642. tion and some unusual properties. Int. J. Biochem. 17:485-494. 39. Scott, A. I., N. J. Stolowich, J. J. Williams, M. D. Gonzales, C. A. 28. Fuhrhop, J. G., and K. M. Smith. 1975. Laboratory methods. In Porphy- Roessner, S. K. Grant, and C. Pichon. 1988. Concerning the active site ofporpho- rins and Metalloporphyrins. K. M. Smith, editor. Elsvier, Amsterdam. 792. bilinogen deaminase. J. Am. Chem. Soc. 110:5898-5900. 29. Brenner, D. A., and J. R. Bloomer. 1980. A fluorimetric assay for measure- 40. Llambias, E. B. C., and A. M. del C Batlle. 1970. Porphyrin biosynthesis in ment of protoporphyrinogen oxidase activity in mammalian tissue. Clin. Chim. soybean callus IX. The porphobilinogen deaminase-uroporphyrinogen III co- Acta. 100:259-266. synthetase. Kinetic studies. Biochim. Biophys. Acta. 220:552-559. 30. Anderson, P. M., and R. J. Desnick. 1980. Purification and properties of 41. Adams, P. A., C. Adams, and D. A. Baldwin. 1986. Ferriprotoporphyrin uroporphyrinogen I synthase from human erythrocytes. Identification of stable IX oxygen activation/insertion reactions: the nature of the interaction between enzyme-substrate intermediates. J. Biol. Chem. 255:1993-1999. ferriprotoporphyrin IX and aniline (substrate). J. Inorg. Biochem. 28:441-454. 31. Corrigall, A. V., P. N. Meissner, and R. E. Kirsch. 1991. Purification of 42. Bonnett, R. 1981. Oxygen activation and tetrapyrroles. In Essays in Bio- human erythrocyte porphobilinogen deaminase. S. Afr. Med. J. 80:294-296. chemistry. P. N. Campbell, and R. D. Marshall, editors. Academic Press, Lon- 32. Bergmeyer, H. U. 1978. Reaction kinetics. In Principles of Enzymatic don. 1-51.

1444 P. Meissner, P. Adams, and R. Kirsch